Microstructures and Mechanical Properties of Shape Memory Alloy Using Pre Mixed TiNi Powders with TiO2 Particles

Microstructures and Mechanical Properties of Shape Memory Alloy Using

Pre-Mixed TiNi Powders with TiO

Particles

Ryoichi Soba

_{, Yukiko Tanabe}

_{, Takayuki Yonezawa}

_{, Junko Umeda}

_{and Katsuyoshi Kondoh}

2_{Joining and Welding Research Institute, Osaka University, Ibaraki 567–0047, Japan}

In this study, microstructural and mechanical properties of the extruded and heat-treated TiNi alloys by sintering the mixture of TiNi pre-mixed powder with titanium dioxide (TiO2) particles were investigated. Pure Ti and pure Ni powder with TiO2 particles were mixed and consolidated by spark plasma sintering (SPS). SPSed TiNi alloy compacts were extruded and heat-treated subsequently. SPSed TiNi alloy compacts had TiNi matrix and Ti4Ni2O phase. Ti4Ni2O phase was formed during SPS by reaction between TiNi matrix and oxygen atoms originated from additive TiO2 particles. Consequently, the heat-treated Ti-50.5 at%Ni alloy using pre-mixed powder with 1.0 vol% TiO2 parti-cles showed a high plateau stress of 630 MPa and a good shape recovery of 79.7% in 8% strain applied. The heat-treated TiNi alloy with 1.0 vol% TiO2 particles revealed the high strength and good shape memory properties. The high strengthening mechanism of the TiNi alloy using pre-mixed powder with TiO2 particles was mainly due to a decrease martensitic transformation temperature by an increase solute Ni content in TiNi matrix after reaction between TiNi and TiO2. [doi:10.2320/matertrans.Y-M2017848]

(Received September 6, 2017; Accepted October 15, 2017; Published December 25, 2017)

Keywords:  TiNi pre-mixed powder, TiO2 particles, shape memory alloy, precipitation, hysteresis

1.  Introduction

Shape memory alloy (SMA) are practically applied owing to excellent properties, such as shape memory effect with re-shape properties when transformed and then heated, super-elastic properties with the recovery of the original shape when unloading and damping characteristics with vibrations attenuated1)_{. As TiNi alloy has excellent superelastic} proper-ties and good corrosion resistance2)_{, it is utilized for} manu-facturing medical devices such as orthodontic wires and medical guidewires. Moreover, in recent years, TiNi alloy is also applied to manufacture stent, which is an implanted medical device in human body and has high risk after being implanted3–5)_{. Stents with tube-shaped and netlike metal} ma-terials are one of devices for the catheter treatment (percuta-neous coronary intervention) and deployed in the vessel ste-nosis or occlusion to promote enough blood flow by expanding and supporting the vessels. Materials plastically deformed such as stainless steel or CoCr alloys are mainly used as in the coronary artery, while TiNi alloys with super-elastic properties are preferably used for the treatment of the artery of lower limb that may transformed by anatomy movement6)_{. This catheter treatment is needed to further} in-vasive procedure such as down-sizing devices. In an effort to maintain the device performance by making diameter small, one of the solutions is to use materials with high strength. Previous studies revealed that TiNi alloy may have high strength and high recovery shape rate by solid-phase sinter-ing of TiNi alloy powder with TiO2 particles7)_{. TiNi alloy} powder produced from homogenized a TiNi ingot is easy to obtain homogenous sintered compact and express mechani-cal/shape-memory properties equal to TiNi alloy castings8)_, although it has a difficulty in arbitrarily changing composi-tion ratio and limited to design materials. TiNi alloy powder

has also economic problems of an increase in processing costs for changing TiNi ingot into powder. In this study, the microstructural and mechanical properties of TiNi alloy with pre-mixed powder of pure Ti powder and pure Ni powder is investigated, which are easy to control TiNi composition ra-tio and inexpensive materials in comparison with TiNi alloy powder. The mechanism of high-strengthened the extruded and heat-treated TiNi alloys by sintering the mixture of TiNi pre-mixed powder with titanium dioxide (TiO2) particles is also investigated.

2. Experimental Procedure

Pure Ti powder with 99.6% purity (TC-450: Toho Technical Service Co., Ltd.) and pure Ti powder with 99.6% purity (SFR-Ni: Nippon Atomized Metal Powders Corporation) were used as a starting material. The results obtained from element analysis of both powder shows in Table 1. Titanium (IV) oxide with 99.5% purity (TiO2, KISHIDA CHEMICAL Co., Ltd.) was used as additive par-ticles. The mixing ratio of Ti powder to Ni power was 50.0 at%Ni, 50.5 at%Ni, 51.0 at%Ni and Ti-52.0 at%Ni, respectively. In the mixing procedure, after weighing pure Ti powder and pure Ni powder according to prescribed mixing ratio, the powder was sealed in the plastic pot and mixed with desktop ball mill (AV-2: Asahi Rika Factory. Ltd). The mixed powder was sealed with ZrO2 ball (YTZ ball: φ10 mm, a powder-to-ball weight ratio of 10:1) to mix Ti powder and Ni powder uniformly. The mixed con-dition was 90 rpm in rotation speed and 10.8 ks in mixed

*1 _{This Paper was Originally Published in Japanese in J. Jpn. Soc. Powder} Powder Metallurgy 64 (2017) 589–594.

*2 _{Corresponding author, E-mail: Ryouichi_Souba@terumo.co.jp}

Table 1 Elemental content analysis of pure Ti and pure Ni powders.

Raw powder mass%

O C Fe Si P

Pure Ti powder 0.22 0.01 0.03 0.01

-Pure Ni powder 0.33 0.01 0.006 - 0.028

time, respectively. In manufacturing of TiNi pre-mixed pow-der added TiO2 particles, after weighing pure Ti, pure Ni powder and TiO2 particles according to prescribed mixing ratio (0 vol%TiO2, 0.5 vol%TiO2 and 1.0 vol%TiO2), the powder was sealed in the plastic pot and mixed with ZrO2 ball (φ10 mm, ball to powder weight ratio of 10:1) to pro-mote the cancellation of TiO2 particles aggregation. The mixed condition was 90 rpm in rotation speed and 20.6 ks in mixed time, respectively. They were consolidated by spark plasma sintering (SPS) (SPS-1030S: SPS SYNTEX) at 1323 K for 3.6 ks in vacuum ( 6 Pa) by applying 40 MPa pressure, and were fabricated to a sintered billet with 36 mm diameter. The dimensions and weight of the sintered billet was measured respectively and the relative density of the sample was calculated. The SPSed billets were pre-heated at 1373 K for 600 s in argon (Ar) gas atmosphere with infrared gold image furnace (RHL-P610C: ULVAC, Inc.), and imme-diately were completely fabricated to powder metallurgy (PM) TiNi alloy rods by hot extrusion (SHP-200-450: Shibayama Kikai Co., Ltd.). Hot extrusion processing condi-tions was that extrusion ratio was 6 (container with φ37 mm inner diameter, die with φ15 mm diameter) and that extru-sion ram speed was 6 mm/s. Moreover, the homogenization heat treatment (retention temperature at 1273 K, 43.2 ks of retention time, furnace cooling) was applied to the extruded bars for homogenization of organization/composition by vacuum furnace (FT-1200-120: FULL-TECH FURNACE Co., Ltd.). Then, the solution heat treatment (retention tem-perature at 1273 K, 3.6 ks of retention time, 3 ℓ/min Ar gas of inlet velocity, water quenching) and the shape memory heat treatment (773 K of retention temperature, 3.6 ks of re-tention time, 3 ℓ/min Ar gas of inlet velocity, water quench-ing) were applied to the rods to precipitate Ti3Ni4 into the TiNi matrix. The structural analysis and composition analy-sis to the obtained sample were employed with field emis-sion type scanning electron microscope and energy disper-sive X-ray spectroscopy (FESEM-EDS JSM-6500F: JOEL Ltd.), electron probe micro analyzer (EPMA, JXA-8530F: JOEL Ltd.) and X-ray diffraction (XRD-6100: Shimadzu Corporation). Tensile tests were performed under a strain rate of 5 ×  10−4_{/s at room temperature (298 ±  2 K) using a} testing machine (Autograph AG-X: Shimadzu Corporation) on tensile bar specimens (a diameter of 3 mm, a gauge length of 15 mm), machined from the extruded rods along the extrusion direction. The hysteresis test was also per-formed under the same conditions and shape recovery rates was measured at 3% and 8% strain applied, respectively.

3. Results and Discussions

3.1  Microstructural and mechanical properties of PM TiNi alloy using pre-mixed powder

The microstructure by SEM of SPSed TiNi using pmixed powder (Ti-50.5 at,%Ni) are shown in Fig. 1. The re-sults exhibited the structure that dispersed compound phase into matrix phase and unreacted pure Ti and pure Ni were not detected. There was no void in the dense structure with 99.9% relative density. Table 2 shows quantitative analysis data of compound phase with EMPA. As 13 at% oxygen at-oms were contained in dispersion phase of a sintered

com-pact and the ratio of Ti to Ni was 2 : 1, it was concluded that this dispersion phase was Ti4Ni2O. As sintering in the sinter-ing process was carried out under vacuum atmosphere, it is conceivable that this Ti4Ni2O was indicated as the oxide that was originated from oxygen atoms contained in the based powder shown in Table 1. Then, SEM observation data at the cross section of as-extruded and homogenization heat treated TiNi alloys are show in Fig. 2. In Ti4Ni2O phase con-taining SPSed TiNi alloy, fine spherical particles with 1.2 µm mean diameter were dispersed in the entire matrix by performing hot extrusion process. Furthermore, as there are no cracks on the interface between Ti4Ni2O phase and ma-trix phase in extruded TiNi alloy, it is conceivable that Ti4Ni2O phase can have high interface consistency to the matrix phase. Before and after homogenization heat treat-ment, significant changes in dispersion regime of Ti4Ni2O particles were not seen. The XRD data of specimens per-formed solution heat treatment and shape memory heat treatment are shown in Fig. 3 and SEM observation data are shown in Fig. 4. In solution heat treated TiNi alloy, diffrac-tion peaks corresponding to Ti3Ni4 on the homogenizadiffrac-tion heat treated TiNi alloy disappeared and needle-like disper-sion were not detected from SEM data. This result showed that needle-like Ti3Ni4 precipitates soluted into TiNi matrix by solution heat treatments. In contrast, in shape memory

Table 2 EPMA point analysis on compounds of SPSed Ti-50.5 at%Ni alloy.

Element Ti Ni O

at% 56.5 30.5 13.0

Fig. 2 SEM observation results of as-extruded (a) and homogenization heat treated (b) Ti-50.5 at%Ni alloys.

heat treatment TiNi alloy, diffraction peaks corresponding to Ti3Ni4 were detected with XRD and the compound precipi-tated fine particles with below 200 nm were observed by SEM observation. Figure 5 shows hysteresis test data of shape memory heat treated PM TiNi alloy using pre-mixed powder (Ti-50.5 at%Ni) and shape memory heat treated PM TiNi alloy using pre-alloyed powder (Ti-51.19 at% Ni) un-der the same condition of manufacturing with extrusion and heat treatment. Shape recovery rates (R) were calculated from following formula and calculated data are presented in Table 3.

R=ST−SR

S_T ×100 (1)

where, ST; applied strain, SR; residual strain. Both ST and SR

are present in plateau region. In the heat treated PM TiNi al-loy using pre-mixed powder, the plateau stress and the end plateau region corresponding in strain were 370 MPa and 4.7%, respectively. By contrast, in the heat treated PM TiNi alloy using pre-alloyed powder which the same heat treat-ment was performed to, the plateau stress and the end pla-teau region corresponding in strain were 289 MPa and 6.1%, respectively. Although the strain of the PM TiNi alloy using pre-mixed powder was 2% lower than one of the PM TiNi alloy using pre-alloyed powder, the plateau stress of the PM

TiNi alloy using pre-mixed powder was 80 MPa higher than one of the PM TiNi alloy using pre-alloyed powder, so the PM TiNi alloy using pre-mixed powder has sufficient me-chanical properties to be used as a SMA. When the stress was unloaded at 3% strain amount in the middle area of pla-teau region, 99.5% shape recovery rate was obtained. This result could be explained that the critical stress of slip im-proved by the matrix phase strengthening action of fine and uniformly precipitated Ti3Ni4 particles, and by suppressing irreversible deformation by the slip9–11)_{. Moreover, when the} PM TiNi alloy using pre-mixed powder was transformed at 8% strain amount in the work hardening region, approxi-mately 60% formation of the material recovered owing to unloading the stress. This shape recovery rate was higher compared to the one of the PM TiNi alloy using pre-alloy powder which the same heat treatment was performed to.

3.2  Microstructures and mechanical properties of PM TiNi alloy using pre-mixed Ti-Ni-TiO2 particles

SEM observation images of TiNi pre-mixed powder (Ti-50.5 at%Ni) with 0.5 vol% and 1.0 vol% TiO2 particles are shown in Fig. 6. TiO2 particles such as aggregation were not identified, and TiO2 particles were found to attach and disperse on the surface of TiNi pre-mixed powder due to mixing process. SEM observation images of SPSed TiNi al-loy using pre-mixed powder are presented in Fig. 7. All the added TiO2 particles were decomposed and disappear, SPSed TiNi alloy using pre-mixed Ti-Ni-TiO2 particles be-came hybrid organization TiNi phase, Ti4Ni2O phase and needle-like Ti3Ni4 phase. As unreacted Ti phase, Ni phase, Ti2Ni and TiNi3 phase as intermediate product, SPSed TiNi alloy using pre-mixed powder both with TiO2 particles and without TiO2 particles can react and disperse absolutely. Microstructures by SEM of solution heat treatment and shape memory heat treated extruded TiNi alloy using pre-mixed powder without TiO2 particles (a) and with TiO2 par-ticles (b) and (c) are shown in Fig. 8. Connection and coars-ening of dispersed Ti4Ni2O particles were found with Fig. 3 XRD analysis results of homogenization heat treated (a), solution

heat treated (b) and shape memory heat treated (c) Ti-50.5 at%Ni alloys.

Fig. 4 SEM observation images of solution heat treated (a) and shape memory heat treated (b) Ti-50.5 at%Ni alloys.

Fig. 5 Stress-strain curves of shape memory heat treated TiNi alloy using Ti-50.5 at%Ni pre-mixed powder (a) and Ti-51.19 at%Ni pre-alloyed powder (b) in 3% and 8% strain applied.

Table 3 Shape recovery rate of shape memory heat treated TiNi alloys.

Shape recovery rate, R (%) ST =  3% ST =  8%

pre-mixed powder 99.5 59.6

increasing TiO2 particles content. It is conceivable that this induced by declining the uniform dispersion when mixing the powder due to the increase in amount of additive TiO2 particles and by producing in the presence of a local uneven distribution. In average grain size of TiNi phase, significant differences were not observed among the addition of the amount of TiO2 particles. Data of tensile test for extruded PM TiNi alloy using pre-mixed powder (mixture ratio of Ti-50.5 to Ti-51.5 at%Ni) without TiO2 particles and with 0.5 vol% and 1.0 vol% TiO2 particles were listed in Table 4. The plateau stress of specimens significantly increased with the amount of additive TiO2 particles. Especially for Ti-50.5 at%Ni alloy with 1.0 vol% TiO2 particles expressed good mechanical properties, with 630 MPa plateau stress, which is approximately 1.7 times as many as the plateau stress (370 MPa) of Ti-50.5 at%Ni alloy materials without TiO2 particles. This plateau stress was excellent compared with the plateau stress of PM TiNi alloy (to 600 MPa) in previous studies12–14)_{. On the other hand, the end plateau} re-gion corresponding in strain decreased with the increase in the additions of the amount of TiO2 particles. In shape re-covery rates, when unloading 3% strain in plateau region over 50.5 at% mixed Ni amount, 100% shape recovery rates were shown not the despite the additions of the amount of TiO2 particle. This suggested that a big influence on entire strain were not exerted by dispersing uniformly in the entire substrate without declining ductility of PM TiNi alloy, al-though Ti4Ni2O particles were brittle compounds. Shape re-covery rates showed over 75% when unloading the stress af-ter the specimen was transformed at 8% strain amount in the work hardening region. By contrast, in materials of 50.0 at% mixed Ni amount, when unloading the stress in plateau re-gion, the strain remained over 1% and the shape recovery rate was 52.3%. It is suggested that residual strain was gen-erated due to occurring slip deformation in plateau region, which caused by decreasing the amount of precipitated Ti3Ni4 and decreasing critical stress for slip in matrix phase as the mixed Ni amount equals to the stoichiometric ratio.

3.3  Strengthening mechanisms of PM TiNi alloy using pre-mixed Ti-Ni-TiO2 particles

Previous section revealed that the increased oxygen atoms by decomposing additive TiO2 particles contributed to gen-erate Ti4Ni2O and that the plateau stress of PM TiNi alloy

Table 4 Modified Ni contents XM-Ni and tensile properties of TiNi alloy composites.

Plateau stress Plateau strain UTS Elongation Shape recovery rate, R (%)

MPa % MPa % ST =  3% ST =  8%

Ti-50.0 at%Ni 245.5 5.3 809.8 14.2 52.3 35.2

Ti-50.0 at%Ni+0.5 vol%TiO2 365.0 4.1 985.0 11.0 98.7 46.9

Ti-50.0 at%Ni+1.0 vol%TiO2 447.8 3.65 1102.3 10.1 98.6 57.5

Ti-50.5 at%Ni 369.7 4.65 1016.5 13.0 99.5 59.6

Ti-50.5 at%Ni+0.5 vol%TiO2 532.2 3.65 1173.0 11.1 100 89.0

Ti-50.5 at%Ni+1.0 vol%TiO2 630.0 3.0 1273.1 9.2 100 79.7

Ti-51.0 at%Ni 542.3 3.6 1331.0 14.0 100 87.2

Ti-51.0 at%Ni+0.5 vol%TiO2 693.0 2.95 1360.1 10.1 100 84.7

Ti-51.0 at%Ni+1.0 vol%TiO2 730.3 2.5 1456.8 9.5 100 81.8

Ti-52.0 at%Ni 812.3 2.5 1620.2 11.6 98.2 87.3

Fig. 7 SEM observation images of SPSed materials using Ti-50.5 at%Ni pre-mixed powders with 1.0 vol% TiO2 particles.

Fig. 8 SEM microstructures of shape memory heat treated Ti-50.5 at%Ni alloys without TiO2 particles (a), with 0.5 vol% TiO2 particles (b) and 1.0 vol% TiO2 particles (c).

was risen with the increase in the amount of additive TiO2 particles without depending on Ni content. From these re-sults, when Ti4Ni2O are generated, it is predicted that the changes of mechanical properties are influenced by dispers-ing Ti4Ni2O particle and by changdispers-ing in the amount of Ni solid solution. Although dispersing Ti4Ni2O particle leads to the pinning of the growth of crystal grain, there weren t sig-nificant differences among the average grain sizes in TiNi phase depending on the amount of generated Ti4Ni2O. It is conceivable that the main strengthening mechanism of TiNi alloy occurs owing to changing in the amount of Ni solid solution in matrix phase. Therefore, the relation between the plateau stress and the amount of Ni solid solution accompa-nying with generating of Ti4Ni2O is considered. A schematic image is presented in Fig. 9, when Ti4Ni2O was generated by the reaction of TiNi including 50 titanium atoms and 50 titanium atoms with one of the TiO2. When one TiO2 mole-cule is added, one titanium atom increases, while eight tita-nium atoms and four nickel atoms are consumed by generat-ing Ti4Ni2O. Taken all together, the reduction ratio of titanium atom to nickel atom was 7：4 and the composition ratio of TiNi phase relatively changed to Ni-rich side. Therefore, the amount of Ni solid solution relatively in-creased by generating Ti4Ni2O described as XM-Ni (at%) can

be calculated from the following equation15)_{, when the} amount of nickel contained in TiNi pre-mixed Ti-Ni-TiO2 particles described as XNi (at%) and the amount of oxygen

contained described as X0 (at%).

X_M-Ni= XNi−2XO

100₋7XO ×100 (2)

Then, the strengthening mechanisms of PM TiNi alloy using pre-mixed powder were considered, when the composition ratio of TiNi phase change to Ni-rich side. The improvement of the plateau stress occurred due to decreasing the marten-sitic transformation temperature accompanying with the in-crease in the amount of Ni solid solution in matrix phase. The relation between stress-induced martensitic transforma-tion (σ) and temperature (T) is represented by the following formula from Clausius-Clapeyron equation16)_.

dσ dT =−

ρ∆S

εt (3)

Herein, ρ, ΔS and εt are a density, a transformation entropy and a transformation strain, respectively. Moreover, when the temperature equals to the martensitic transformation start temperature, the plateau stress is calculated by following formula, as σ equals to zero.

σP =ρ∆S

εt (MS −

T) (4)

In this study, when ρΔS/εt equals to −3.03 MPa/K17,18), an increased amount of the plateau stress per unit the amount of Ni solid solution is 256 MPa/at%Ni calculated by formula (4). The relation between the amount of Ni solid solution calculated by formula (2) and the plateau stress listed in Table 4 is shown in Fig. 10. The linear relationship between the plateau stress and the amount of Ni solid solution exists described as the formula (5) and the slope of the formula is 266 MPa/at%Ni, despite the additions of the amount of TiO2 particles. This value coincides with the rate of 256 MPa/at%Ni increase of the plateau stress calculated by the above theoretical formula (4). The increase in the plateau stress could be explained by decreasing the martensitic transformation temperature associated with rising the amount of Ni solid solution in matrix phase, contributed to generate Ti4Ni2O.

σp=265.66 XM-Ni−13300 (5)

4.  Conclusion

In this study, powder metallurgy TiNi alloy using mixed powder and powder metallurgy TiNi alloy using pre-mixed powder with TiO2 particles were manufactured in an effort to be examined for high strength and strengthening mechanism of powder metallurgy TiNi alloy. Microstructural analysis, mechanical properties were per-formed for each TiNi alloy. The result revealed that the com-position ratio of TiNi matrix phase could be controled indi-rectly by adding TiNi pre-mixed powder to TiO2 particles, allowing achieve of high strength/high recovery rate of TiNi alloys. This study provided knowledge as following.

(1) The SPSed TiNi alloys manufactured by pure Ti and Ni pre-mixed powder exhibited the structure that disperses Ti4Ni2O phase originated from oxygen atoms in the based powder. The powder metallurgy Ti-50.5 at%Ni alloy using pre-mixed powder showed 370 MPa plateau stress and 99.5% shape recovery rate in the end plateau region, whose values were available as shape memory alloys.

(2) The plateau stress of the powder metallurgy Ti-50.5 at%Ni alloy using pre-mixed powder with TiO2

parti-Fig. 9 Schematic image of Ti4Ni2O formation by reaction TiO2 with TiNi.

cles significantly increased with increase in the additions of the amount of TiO2 particles. Ti-50.5 at%Ni alloy with 1.0 vol% TiO2 particles expressed good mechanical proper-ties with 630 MPa plateau stress, which is approximately 1.7 times as many as the plateau stress of Ti-50.5 at%Ni alloy without TiO2 particles.

(3) Strengthening mechanisms of the plateau stress was mainly caused by declining the martensitic transformation temperature due to the increase in the amount of Ni solid solution. The plateau stress was treated as a linear approxi-mation proportional to the amount of converted Ni solid solution with the amount of Ni solid solution accompanying with generating Ti4Ni2O phase considered.

Acknowledgments

A part of this study was financially supported by strategic promotion of innovative research and development program promoted by Japan Science and Technology Agency (JST), Japan.

REFERENCES

1) S. Miyazaki, T. Sakuma and T. Shibuya: Applications of Shape

Memory Alloys, (CMC Publishing, 2001).

2) D.J. Wever, A.G. Veldhuizen, J. de Vries, H.J. Busscher, D.R.A. Uges and J.R. van Horn: Biomaterials 19 (1998) 761–769.

3) T. Duerig, A. Pelton and D. Stockel: Mater. Sci. Eng. A 273–275 (1999) 149–160.

4) J.K. Allafi, B.A. Ahmadi and M. Zare: Mater. Sci. Eng. C 30 (2010) 1112–1117.

5) P. Filip, J. Lausmaa, J. Musialek and K. Mazanec: Biomaterials 22 (2001) 2131–2138.

6) N.B. Morgan: Mater. Sci. Eng. A 378 (2004) 16–23.

7) T. Yonezawa, T. Yoshimura, J. Umeda, K. Kondoh and R. Souba:

Smart Processing Society for Materials 1 (2012) 288–292.

8) T. Koyari, K. Hatano, H. Katoh, S. Miura, M. Tokizane: Proceedings of 1991 Society of Materials Science Japan, 40 (1991) pp. 390–392. 9) K. Otsuka and X. Ren: Prog. Mater. Sci. 50 (2005) 511–678.

10) H. Zhao, C.Q. Liang, J.T. Liu, Y.X. Tong, F. Chen, B. Tian, L. Li and Y.F. Zheng: J. Mater. Eng. Perform. 21 (2012) 2566–2571.

11) M. Nishida, T. Hara, T. Ohba, K. Yamaguchi, K. Tanaka and K. Yamauchi: Mater. Trans. 44 (2003) 2631–2636.

12) K. Funami, Y. Sekiguchi and H. Funakubo: J. Jpn. Inst. Metals 48 (1984) 1113–1119.

13) M. Bram, A.A. Khanlou, A. Heckmann, B. Fuchs, H.P. Buchkremer and D. Stöver: Mater. Sci. Eng. A 337 (2002) 254–263.

14) E.C. Yen and K.S. Hwang: Mater. Sci. Eng. A 528 (2011) 5296–5305.

15) Y. Shugo, S. Hanada and T. Honma: Bulletin of the Research Institute of Mineral Dressing and Metallurgy 41 (1985) 23–34.

16) Y. Liu and H. Yang: Smart Mater. Struct. 16 (2007) S22–S27.

17) K. Otsuka and X. Ren: Prog. Mater. Sci. 50 (2005) 511–678.